Use of spatial analysis techniques to identify statistically significant crash hots spots in metropolitan Melbourne

Traditional statistical techniques have limitations in analysing crashes as these techniques assume spatial independence and stationarity. Crashes break these assumptions as they tend to cluster at specific locations (spatial dependency) and vary from one location to another (non-stationarity). Several spatial statistical methods were used to examine crash clustering in metropolitan Melbourne, including the Getis-Ord Gi* method which identified statistically significant crash clusters. Using this method, the degree, location and extent of clustering were found to vary for different crash categories, with fatal crashes exhibiting the lowest level of clustering and bicycle crashes exhibiting the highest level of clustering.